Inferring Depression and Its Semantic Underpinnings from Simple Lexical Choices

Table 2

Study 2: classification performance metrics.

Model	PCs	Accuracy		95% CI				ROC AUC		Sensitivity		Specificity
		Train	Test	Train	Test	Train	Test	Train	Test	Train	Test	Train	Test
mDCT	10	.59	.60	[.54, .64]	[.53, .68]	<.001	<.01	.58	.58	.57	.55	.61	.66
mDCT+Demo	24	.63	.62	[.58, .68]	[.55, .70]	<.001	<.001	.70	.62	.66	.60	.60	.65
mGenderAge	—	.59	.53	[.54, .64]	[.45, .61]	<.001	.25	.61	.60	.61	.58	.58	.48
mRandomBaseline	8	.55	.55	[.50, .60]	[.47, .63]	<.05	.10	.59	.52	.57	.59	.54	.51

The best performing model is highlighted in bold.